UMBC Meyerhoff Scholars Program

Permanent URI for this collectionhttp://hdl.handle.net/11603/14258

The Meyerhoff Scholars Program has been at the forefront of efforts to increase diversity among future leaders in science, engineering, and related fields. The UMBC Meyerhoff family is now more than 1400 strong with over 1100 alumni across the nation, which includes over 300 students enrolled in graduate and professional programs.

The program’s success is built on the premise that, among like-minded students who work closely together, positive energy is contagious. By assembling such a high concentration of high-achieving students in a tightly knit learning community, students continually inspire one another to do more and better.

The program has been recognized by the National Science Foundation and The New York Times as a national model. Scores of representatives from federal agencies, campuses, and corporations across the country have visited UMBC’s campus to learn more about the program’s success. The College Board’s National Task Force on Minority High Achievement praised the Meyerhoff Scholars Program as an example that could provide broader educational lessons.

Browse

Now showing 1 - 12 of 12

Neural filtering for Neural Network-based Models of Dynamic Systems
(2024-09-20) Oveissi, Parham; Rozario, Turibius; Goel, Ankit
The application of neural networks in modeling dynamic systems has become prominent due to their ability to estimate complex nonlinear functions. Despite their effectiveness, neural networks face challenges in long-term predictions, where the prediction error diverges over time, thus degrading their accuracy. This paper presents a neural filter to enhance the accuracy of long-term state predictions of neural network-based models of dynamic systems. Motivated by the extended Kalman filter, the neural filter combines the neural network state predictions with the measurements from the physical system to improve the estimated state's accuracy. The neural filter's improvements in prediction accuracy are demonstrated through applications to four nonlinear dynamical systems. Numerical experiments show that the neural filter significantly improves prediction accuracy and bounds the state estimate covariance, outperforming the neural network predictions.
Visualizing 3-Dimensional Ptychography Data
(2024-08) Stanford, V. Vianne; Bhatt, Lopa; Muller, David
Ruddlesden-Popper Phases (RP) are a method of growing materials to induce specific properties in the material. In nickelates, the proper RP domain gives superconducting properties to the material. Unfortunately, measuring the distribution of RP domains in a sample in a statistically significant area is tedious and time consuming. To aid in this endeavor, we modified a previously created RP-domain tracking code² ̓ ³ to allow for analysis of ptychography data and to add 3-dimensional visualization capabilities.
Impact of Confinement within a Hydrogel Mesh on Protein Thermodynamic Stability and Aggregation Kinetics
(ACS, 2024-01-26) Ghassemi, Zahra; Leach, Jennie B.
Though protein stability and aggregation have been well characterized in dilute solutions, the influence of a confining environment that exists (e.g., in intercellular and tissue spaces and therapeutic formulations) on the protein structure is largely unknown. Herein, the effects of confinement on stability and aggregation were explored for proteins of different sizes, stability, and hydrophobicity when encapsulated in hydrophilic poly(ethylene glycol) hydrogels. Denaturation curves show linear correlations between confinement size (mesh size) and thermodynamic stability, i.e., unfolding free energy and surface area accessible for solvation (represented by m-value). Two clusters of protein types are identifiable from these correlations; the clusters are defined by differences in protein stability, surface area, and aggregation propensity. Proteins with higher stability, larger surface area, and lower aggregation propensity (e.g., lysozyme and myoglobin) are less affected by the confinement imposed by mesh size than proteins with lower stability, smaller surface area, and higher aggregation propensity (e.g., growth hormone and aldehyde dehydrogenase). According to aggregation kinetics measured by thioflavin T fluorescence, confinement in smaller mesh sizes resulted in slower aggregation rates than that in larger mesh sizes. Compared to that in buffer solution, the confinement of a hydrophobic protein (e.g., human insulin) in the hydrogels accelerates protein aggregation. Conversely, the confinement of a hydrophilic protein (e.g., human amylin) in the hydrogels decelerates or prevents aggregation, with the rates of aggregation inversely proportional to mesh size. These findings provide new insights into protein conformational stability in confined microenvironments relevant to various cellular, tissue, and therapeutics scenarios.
Matrix-Based Representations and Gradient-Free Algorithms for Neural Network Training
(IEEE) Rozario, Turibius; Oveissi, Parham; Goel, Ankit
This paper presents a compact, matrix-based representation of neural networks. Although neural networks are often understood pictorially as interconnected neurons, they are fundamentally mathematical nonlinear functions constructed by composing several vector-valued functions. Using basic results from linear algebra, we represent neural networks as an alternating sequence of linear maps and scalar nonlinear functions, known as activation functions. The training of neural networks involves minimizing a cost function, which typically requires the computation of a gradient. By applying basic multivariable calculus, we show that the cost gradient is also a function composed of a sequence of linear maps and nonlinear functions. In addition to the analytical gradient computation, we explore two gradient-free training methods. We compare these three training methods in terms of convergence rate and prediction accuracy, demonstrating the potential advantages of gradient-free approaches.
Assessing departmental readiness to support minoritized faculty
(Wiley, 2024-02-27) Carter-Veale, Wendy Y.; Cresiski, Robin H.; Sharp, Gwen; Lankford, Jordan D.; Ugarte, Fadel
Though increasing numbers of racially and ethnically minoritized (REM) individuals earn PhDs and national initiatives focus on faculty diversity, challenges persist in recruiting, hiring, and retaining REM faculty. While a pervasive issue nationally, the literature predominantly focuses on faculty diversity at research-intensive institutions. This exploratory case study pilots a readiness instrument to evaluate the commitment and willingness of a biomedical department at a primarily undergraduate institution to embrace faculty diversity before initiating a postdoctoral faculty conversion program. We introduce the Community Readiness Model (CRM) into an academic context, offering academic departments a robust framework and tool to evaluate readiness and capacity to recruit, retain, and support REM faculty. Practical Takeaways Academic departments can be conceived of as a type of community. The adapted Department Readiness Tool can be a valuable method of evaluating a department's readiness to support the success of underrepresented minority faculty. Departments may score highly on some areas of readiness but relatively low on others, which provides insight into where time and resources should be invested to improve readiness.
From Exclusion to Opportunity
(Georgetown Center on Poverty and Inequality, 2022-08-24) PALACIOS, VINCENT; TATUM, LAURA; COOPER, NATALIA; ANEJA, SIDDHARTHA
A four-year postsecondary degree offers opportunities for a higher income and upward economic mobility. However, postsecondary education—historically inaccessible to people of color and women—also plays a key role in reproducing and amplifying societal inequities by sorting students into specialized fields of study by race and gender, contributing to a segregated labor force. This report examines the link between postsecondary field of study and labor market segregation using an original quantitative analysis. This report presents four principles and corresponding recommendations that postsecondary institutions and policymakers can use to reduce racial and gender segregation across fields of study, increase degree attainment, and ultimately, ameliorate labor market segregation.
Catherine Clarke Fenselau
Catherine Clarke Fenselau (born 15 April 1939) is an American scientist who was the first trained mass spectrometrist on the faculty of an American medical school; she joined Johns Hopkins School of Medicine in 1968. She specializes in biomedical applications of mass spectrometry.[4] She has been recognized as an outstanding scientist in the field of bioanalytical chemistry because of her work using mass spectrometry to study biomolecules.
Keeping Black students in STEM
(Proceedings of the National Academy of Sciences, 2021-06-08) Suran, Melissa
Activity Dependent Modulation of Granule Cell Survival in the Accessory Olfactory Bulb at Puberty
(Frontiers Media, 2020-05-23) Oboti, Livio; Trova, Sara; Schellino, Roberta; Marraudino, Marilena; Harris, Natalie R.; Abiona, Olubukola M.; Stampar, Mojca; Lin, Weihong; Peretto, Paolo
The vomeronasal system (VNS) is specialized in the detection of salient chemical cues triggering social and neuroendocrine responses. Such responses are not always stereotyped, instead, they vary depending on age, sex, and reproductive state, yet the mechanisms underlying this variability are unclear. Here, by analyzing neuronal survival in the first processing nucleus of the VNS, namely the accessory olfactory bulb (AOB), through multiple bromodeoxyuridine birthdating protocols, we show that exposure of female mice to male soiled bedding material affects the integration of newborn granule interneurons mainly after puberty. This effect is induced by urine compounds produced by mature males, as bedding soiled by younger males was ineffective. The granule cell increase induced by mature male odor exposure is not prevented by pre-pubertal ovariectomy, indicating a lesser role of circulating estrogens in this plasticity. Interestingly, the intake of adult male urine-derived cues by the female vomeronasal organ increases during puberty, suggesting a direct correlation between sensory activity and AOB neuronal plasticity. Thus, as odor exposure increases the responses of newly born cells to the experienced stimuli, the addition of new GABAergic inhibitory cells to the AOB might contribute to the shaping of vomeronasal processing of male cues after puberty. Consistently, only after puberty, female mice are capable to discriminate individual male odors through the VNS.
Design of materials for IR detectors using high Z elements for high energy radiation environment
(SPIE, 2019-05-13) Saraf, Sonali; Cooper, Christopher; Su, Ching Hua; Choa, Fow-Sen; Arnold, Bradley; Cullum, Brian; Sachs, David; DiPaula, Philip; Singh, N. B.
There is a strong need for rad hard and high operating temperature IR detectors for space environment. Heavy metal Selenides (high Z and large density) have been investigated for more than half century for high operating temperature mid wave infrared (MWIR) applications. Most of the efforts have been devoted to make detector arrays on high-resistivity Si substrates for operating wavelengths in the 1.5 to 5.0 μm region using physical vapor transport grown poly crystalline materials. For most of the biological spectral and imaging applications, short wave infrared (SWIR) detectors have shown better performance. Recent growth materials have shown variation in morphology with slight change in growth conditions and hence variation in performance parameters such as bandgap, mobility and resistivity from sample to sample. We have performed growth and optical characterization of binary materials CdSe-PbSe to determine the suitability for IR detector. We have determined bandgap using several theoretical models for different morphologies observed during growth on silicon wafers.
4 pillars of college success in science
(TED Conferences, LLC, 2013-02) Hrabowski, Freeman
At age 12, Freeman Hrabowski marched with Martin Luther King. Now he's president of the University of Maryland, Baltimore County (UMBC), where he works to create an environment that helps under-represented students -- specifically African-American, Latino and low-income learners -- get degrees in math and science. He shares the four pillars of UMBC's approach.
Replicating Meyerhoff for inclusive excellence in STEM
(American Association for the Advancement of Science, 2019-04-26) Sto. Domingo, Mariano R.; Sharp, Starlette; Freeman, Amy; Freeman Jr., Thomas; Freeman, Thomas, Jr.; Harmon, Keith; Wiggs, Mitsue; Sathy, Viji; Panter, Abigail T.; Oseguera, Leticia; Sun, Shuyan; Williams, Mary Elizabeth; Templeton, Joseph; Folt, Carol L.; Barron, Eric J.; Hrabowski III, Freeman A.; Maton, Kenneth I.; Crimmins, Michael; Fisher, Charles R.; Summers, Michael F.

Browse

Recent Submissions